Exploratory Data Analysis using Artificial Neural Networks

Abstract

Data analysis helps travel organizations to provide better recommendations for investing in their future trips based on its business and personal trips. This paper presents the basic concepts, various types and levels of data analysis, predictive modeling techniques and appropriate performance measures. There are basically three types of algorithms for predicting such as linear regression (machine learning model), analysis of Variance (statistical model) and artificial neural network (machine learning model). Data Analysis is being used in many fields such as health care, manufacturing, information technology and so on. A travel dataset provided by the uber in Kaggle is used to study the performance of chosen predicting algorithms. The primarymethodology behind this study is to analyze and find the accuracy of the most frequent category of trip among all trips taken by a customer in a region using data analysis. The parameters which are taken into consideration are category, purpose, total distance and speed of the travel. The results of precision, recall, f1 score, Area Under Curve (AUC) and Receiver Operating Characteristic Curve (ROC) are evident that the Artificial neural network (ANN) based prediction is comparatively higher than other algorithms.